source("R/utils.R")
source("R/mcmc.R")
source("R/functions.R")
source("R/monitornew.R")
load_pkgs()
zmargin <- theme(panel.spacing = grid::unit(0, "lines"))
theme_set(theme_bw())
library(targets)
tar_load(traceplot_ag_mcmc0)
print(traceplot_ag_mcmc0)
tar_load(ag_mcmc0)
aa <- do.call(abind, c(ag_mcmc0, list(along=3)))
aa2 <- aperm(aa,c(1,3,2), resize=TRUE)
monitor(aa2)
## Inference for the input samples (8 chains: each with iter = 8000; warmup = 0):
##
## Q5 Q50 Q95 Mean SD Rhat Bulk_ESS Tail_ESS
## loss.sc 2.93 3.23 3.51 3.22 0.18 1.00 20667 33530
## loss.pc 4.26 4.73 5.13 4.71 0.26 1.00 15128 24480
## loss.ag_pc0_sc0 2.47 4.07 5.54 4.05 0.93 1.00 2295 5106
## gain.sc 4.91 5.13 5.34 5.13 0.13 1.00 22464 35882
## loss.ag_pc0_sc1 2.30 4.36 6.90 4.45 1.38 1.01 1137 1870
## gain.pc 3.55 3.93 4.27 3.92 0.22 1.00 15862 29335
## loss.ag_pc1_sc0 3.81 4.58 5.20 4.55 0.42 1.00 8589 16730
## loss.ag_pc1_sc1 2.12 3.59 4.78 3.54 0.81 1.00 2699 5605
## gain.ag_pc0_sc0 1.58 2.52 3.26 2.48 0.52 1.00 5264 11605
## gain.ag_pc0_sc1 0.27 1.63 2.91 1.62 0.81 1.00 1821 2496
## gain.ag_pc1_sc0 3.52 4.23 4.82 4.21 0.40 1.00 9868 16006
## gain.ag_pc1_sc1 0.93 2.68 4.27 2.65 1.01 1.01 1696 3758
##
## For each parameter, Bulk_ESS and Tail_ESS are crude measures of
## effective sample size for bulk and tail quantities respectively (an ESS > 100
## per chain is considered good), and Rhat is the potential scale reduction
## factor on rank normalized split chains (at convergence, Rhat <= 1.01).
bp2 <- ggpairs(as.data.frame(lump.mcmc.list(ag_mcmc0)), progress=FALSE,
lower=list(continuous=function(...) my_mcmc(..., show_prior=FALSE)),
upper=list(continuous=function(...) my_mcmc(..., geom="density", show_prior=FALSE))) +
zmargin
bp2_time <- system.time(print(bp2))
Contour levels are: 50%, 80% 90%, 95% (largest) highest posterior density regions.
tar_load(traceplot_ag_mcmc_tb)
print(traceplot_ag_mcmc_tb)
tar_load(ag_mcmc_tb)
aa <- do.call(abind, c(ag_mcmc_tb, list(along=3)))
aa2 <- aperm(aa,c(1,3,2), resize=TRUE)
monitor(aa2)
## Inference for the input samples (8 chains: each with iter = 8000; warmup = 0):
##
## Q5 Q50 Q95 Mean SD Rhat Bulk_ESS Tail_ESS
## loss.sc 3.06 3.35 3.62 3.35 0.17 1.00 20376 33951
## loss.pc 4.45 4.87 5.24 4.86 0.24 1.00 14007 26347
## loss.ag_pc0_sc0 2.45 4.09 5.58 4.06 0.95 1.01 1976 4736
## gain.sc 5.03 5.23 5.43 5.23 0.12 1.00 22759 35775
## loss.ag_pc0_sc1 2.31 4.31 6.84 4.40 1.36 1.00 1127 1870
## gain.pc 3.60 3.97 4.31 3.96 0.22 1.00 20518 33001
## loss.ag_pc1_sc0 4.07 4.75 5.31 4.73 0.38 1.00 11503 16782
## loss.ag_pc1_sc1 2.25 3.73 4.94 3.68 0.82 1.00 2658 4610
## gain.ag_pc0_sc0 1.39 2.39 3.16 2.35 0.54 1.00 4853 8425
## gain.ag_pc0_sc1 0.28 1.65 2.87 1.63 0.79 1.00 1919 2646
## gain.ag_pc1_sc0 4.20 4.79 5.26 4.76 0.34 1.00 10934 11859
## gain.ag_pc1_sc1 1.02 2.80 4.37 2.76 1.02 1.00 1723 3571
##
## For each parameter, Bulk_ESS and Tail_ESS are crude measures of
## effective sample size for bulk and tail quantities respectively (an ESS > 100
## per chain is considered good), and Rhat is the potential scale reduction
## factor on rank normalized split chains (at convergence, Rhat <= 1.01).
bp3 <- ggpairs(as.data.frame(lump.mcmc.list(ag_mcmc_tb)), progress=FALSE,
lower=list(continuous=function(...) my_mcmc(..., show_prior=FALSE)),
upper=list(continuous=function(...) my_mcmc(..., geom="density", show_prior=FALSE))) +
zmargin
bp3_time <- system.time(print(bp3))